Chapter 2

IDEALISM AND REALISM IN EARLY ANALYTIC PHILOSPHOPHY

Wittgenstein’s Tractatus Logico-Philosophicus

Kant began a new era in German philosophy, but it was not until the end of the nineteenth century that idealism had a profound affect on British philosophy.  The golden age of British idealism was led by T. H. Green, who put forward a Kantian critique of empiricism.  He argued that certain formal conceptions, or ‘categories’, were necessary conditions for the possibility of experience.  By conceiving of the real as totally independent of the mind, the empiricist makes it inaccessible to knowledge and thought, and thus makes it the ‘unmeaning, the empty, of which nothing can be said’[1].  But like Schopenhauer, Green thought that Kant had not held firmly to his idealist principles.  For Green, this failure was not only manifest in Kant’s acceptance of ‘the given’, but also in Kant’s talk of the thing-in-itself, the noumenal world beyond the conditions of experience.  His rejection of these aspects leaves Green with a philosophy that could be described as ‘Hegelian’.[2]

Central to the ‘absolute idealism’ made popular by philosophers such as Green and Bradley was adherence to some form of the coherence theory of truth.  Propositions, it was claimed, cannot be assigned truth-values independently, but must be considered as part of ‘an organic system’.  And in contrast to a correspondence theory of truth, a coherence theory claims that propositions cannot be considered as only externally related to facts.  Rather, both facts and propositions are considered as abstractions from judgement.

As a follower of Bradley, Bertrand Russell was himself an idealist for a period, but as he became more involved with the philosophy of logic, he turned away from idealism.  His initial ‘refutation’ concentrated on attacking the coherence theory of truth[3].  One problem emphasised by Russell was that, since a proposition cannot be considered in isolation from the system of propositions, it could not be considered absolutely true.  Propositions were said to be ‘more or less true’, a view that Russell found to be absurd, for it was self-refuting.  The very suggestion of a coherence theory turns out not to be absolutely true, but only more or less true.  Thus Russell insisted that beliefs were true independently of each other.  Furthermore, they were made true by mind independent facts to which they were externally related.

The traditional theory of truth more often favoured by empiricists is the correspondence theory, according to which a proposition is true if and only if it corresponds to the facts.  While Russell initially rejected this alternative, he later formulated his views in terms of correspondence.  In ‘The Nature of Truth and Falsehood’ (1910) he uses the term ‘correspondence’ somewhat cautiously, but in The Problems of Philosophy he says that ‘a belief is true when it corresponds to a certain associated complex, and false when it does not.’[4]  Above all else, this way of putting the matter allows him to emphasise his main point that while beliefs depend on minds for their existence, they do not depend on minds for their truth.[5]

It was soon after his rejection of idealism that Russell met and taught Ludwig Wittgenstein.  In the opening sections of Wittgenstein’s Tractatus one can see Russell’s influence in an apparent commitment to something like Russell’s realism.  “The world is the totality of facts,” states Wittgenstein, and it is these facts that make a proposition true or false.  The theory of meaning that emerges from what follows involves some elements that one might find in a correspondence theory of truth.  Wittgenstein does not define truth as correspondence with the facts, but on his theory of meaning, propositions are pictures of facts, and the elements of the picture correspond to the elements of the fact it represents.  A picture either agrees or disagrees with reality[6], and its truth consists in the agreement of its sense with reality[7].  Furthermore, elementary propositions are independent of one another, in that their truth-values do not depend on one another.

This is by no means the only influence Russell had on Wittgenstein’s early work[8].  Wittgenstein inherited from Russell and Frege a certain conception of philosophy that placed the philosophy of language at the centre, with the logical analysis of both language and mathematics providing the foundations.  Wittgenstein’s Tractatus Logico-Philosophicus is largely a work in the philosophy of language and logic, and much of the detail can be attributed to the influence of Russell and Frege, or to a reaction to their work.  It would be a mistake however, to see the Tractatus solely in relation to the philosophies of Russell and Frege.  While Wittgenstein inherited the conception of philosophy with linguistic analysis at its centre, the nature of its importance is given a distinctive character by Wittgenstein.  And this character, I will argue, is both Kantian and mystical.

It is often remarked that the Tractatus was influenced by Kant ‘through Schopenhauer’, who he was known to have read extensively.  Certainly Schopenhauer’s influence can be seen throughout the Notebooks and in the section of the Tractatus that deals with solipsism.  But I think it is implausible that the Kantian character of the Tractatus derives exclusively from Schopenhauer’s The World as Will and Representation[9].  Perhaps Wittgenstein had already read Kant before he wrote the Tractatus[10], or perhaps he learnt of his work through discussion, possibly with Russell.  Either way, the general approach of the Tractatus has more in common with Kant’s work than Schopenhauer’s.  It is primarily concerned with the limits of the world and the bounds of metaphysics.  The Tractatus argues that the world is limited by the conditions, not of experience, however, but of representation.  The parallel with Kant lies in the combination of this idealism with the conceptual realism that supports it.  According to Wittgenstein, the subject matter of philosophy is just those conditions of representation, but those conditions are not subject to themselves.  Strictly speaking then, they are not describable at all.  They make themselves manifest.  And therein lies the mysticism of the Tractatus.  The combination of conceptual realism and conceptual idealism in Wittgenstein’s philosophy results in a third position: quietism.

2.1 The Metaphysics of the Tractatus

The Tractatus Logico-Philosophicus is based on some fundamental assumptions about the connection between logic, philosophy and the world.  Wittgenstein presents a picture of the logical form of the world and language — a form that the world and language must share if language (and thought) is to be possible at all.  The picture Wittgenstein presents at the beginning of the Tractatus goes something like as follows.

The world is the totality of facts.  Facts are constituted by atomic facts and atomic facts consist of objects standing in relation to one another.  Analogously, descriptive language is the totality of propositions, propositions are constituted by (are truth functions of) elementary propositions and elementary propositions consist of names standing in relations to each other.

The link between language and the world is made by two semantic relations: naming and picturing.  Both are forms of representation: Names name objects and propositions picture possible facts.  The elementary propositions picture possible atomic facts:

4.311 One name stands for one thing, another for another thing, and they are combined with one another.  In this way the whole group — like a tableau vivant — presents a state of affairs.

If the picture presented by elementary propositions agrees with reality, then that picture (and thus the proposition) is true.  It is false if the possible atomic fact that it represents does not exist in the actual world.

All this metaphysical mechanism underlies ordinary language.  The ‘Names’ are not the proper names or definite descriptions of ordinary language, and ‘Objects’ are not the ordinary objects[11].  Rather, they are to be found at the theoretical limits of analysis.  A fully analysed sentence of English is supposed to consist of groups of simple Names concatenated, which in turn are related to one another by some truth functional operator.  Thus we have the three levels of analysis shown in figure 1:

Fig.  1: Levels of analysis in the metaphysics of the Tractatus.

The bottom two levels are theoretical requirements to a theory of meaning.  Objects, for instance, are ‘simples’.  They are rather puzzling theoretical entities, and Wittgenstein, it seems, never made up his mind what they were besides specifying some of their properties in a rather abstract manner[12].  Neither did he think it was necessary to specify them in more detail in order to see that they exist (subsist).  They were required on transcendental grounds.

2.2 The Realism of the Tractatus: Objects as ‘the Given’

According to Wittgenstein, there subsist simple objects that cannot be broken down into constituent parts:

2.02 Objects are simple.

2.0201 Every statement about complexes can be resolved into a statement about their constituents and into the propositions that describe the complexes completely.

Object are absolutely simple so that we can (at least in theory) reduce statements about ordinary objects down to elementary propositions that relate to the world.  To say that Objects are simples that make up the substance of the world is to say that this theoretical limit of analysis exists, and that it corresponds in some primitive way to the structure of the world.  Wittgenstein’s point is that it must exist if language is to picture the world.  The world and language must share a structure: they must share their logical form.  The requirement that objects are simple and unalterable (2.026 – 2.0271) is the requirement that they cannot be described by means of a further (contingent) proposition.  That words stand for objects must be a necessary truth.  It cannot depend on whether another (empirical) proposition is true, for then it would be a contingent matter whether or not it had sense.  And this cannot be the case if language is to depict the world.  That propositions already have sense is a necessary condition on them being true or false:

2.021 Objects make up the substance of the world.  That is why they cannot be composite.

2.0211 If the world had no substance, then whether a proposition had sense would depend on whether another proposition was true.

2.0212 In that case we could not sketch any picture of the world (true or false).

This point can also be made in terms of possibility.  What is possible is equivalent, for Wittgenstein, with what is thinkable (what is imaginable, 2.022), and this is equivalent to what is representable (that which we can picture to ourselves, 3.001).  The conditions for representation are thus given for all possible worlds, so these conditions cannot be given by what happens to be the case in this or that particular possible world.  These conditions, which include the requirement that there are simple Objects, are also the conditions that a proposition has sense.  Thus, whether there are objects cannot depend on any empirical proposition being true or false.  It is a necessary condition for us to be able to describe the world (2.211):

2.022 It is obvious that an imagined world, however different it may be from the real one, must have something — a form — in common with it.

2.023 Objects are just what constitute this unalterable form.

The requirement that there are simple objects could also be called a ‘transcendental’ condition.  While there is some controversy concerning what properly constitutes a ‘transcendental argument’, a rough characterisation will be sufficient for our purposes.  Kant uses the term for anti-sceptical arguments for (synthetic) a priori claims established on the basis that their truth is necessary for the possibility of experience.  But more generally, a transcendental argument is any argument where a certain phenomenon p (or the structure of this phenomenon) is argued to have a set of conceptually necessary conditions S[13], thus establishing the ‘transcendental validity’ of the concepts included in those conditions. Whilst Kant was interested in the conditions of knowledge, the Wittgenstein of the Tractatus was interested in the conditions for describing the world.  These necessary conditions included the idea that the sense of propositions must be independent of whether they are true or false (2.22 What a picture represents it represent independently of its truth or falsity, by means of its pictorial form.)  There must be contact between language and the world that is prior to the truth or falsity of any proposition about the world.  This contact he characterises with the relations of naming and picturing, relations that presuppose a certain logical form that is shared by language and the world.

The Objects of the Tractatus play a role that is in some ways analogous to the role played by ‘the given’ in Kant’s Transcendental Idealism.  According to Kant, the faculty of sensibility is distinct from the faculty of the understanding.  It is through the sensibility that the mind has a ‘receptivity’ to the given, and through the understanding that they are thought about.  Both are required for meaningful thought.  “Thoughts without content are empty, intuitions without concepts are blind”.[14]  A thought without content is a thought that does not take in elements of the given.  Such thoughts are merely representations, produced by the understanding, but without real application.  The psychological idiom in which this idea is cast by Kant is completely missing from Wittgenstein’s philosophy of logic and language, but a similar idea is central to his thinking.  Any thoughts that do not picture possible states of affairs are senseless precisely because they do not picture combinations of objects.

Now this idea that Kant and Wittgenstein share contains a problem.  The thought that ‘thoughts without content are empty’ is devoid of any specific reference to the given.  It is without content, and therefore empty.  Similarly, the thoughts about Objects in general are presumably not combinations of names that refer to specific Objects.

Both the given and tractarian Objects are transcendentally ‘real’ in that they are independent of the subjective conditions of thought and representation.  The given cannot be considered to have conceptual form since it provides the raw data on which the understanding grounds its lowest level conceptualisations.  Elements of the given form part of the very conditions for meaningful (non-empty) thought.  Likewise, tractarian Objects are not part of what can be described, but are part of the conditions for being able to describe anything.  But then how can we speak of them at all?[15]  Wittgenstein considered his radical solution to this problem as the central point of his early philosophy.  Strictly speaking, we cannot speak of them (and Wittgenstein acknowledges that his own propositions are nonsense, 6.54), but the fact that we can speak at all, shows that they exist.  They make themselves manifest.

2.3 Showing and Saying

After receiving Russell’s comments on his work, Wittgenstein responded by writing that Russell had not understood his ‘main contention, to which the whole business of logical propositions is only corollary.’  He goes on:

The main point is the theory of what can be expressed (gesagt) by propositions — i.e. by language (and what comes to the same, what can be thought) and what cannot be expressed by propositions, but only shown (gezeigt); which, I believe, is the cardinal problem of philosophy…[16]

This ‘cardinal problem’ is related to the task that Wittgenstein claims to be addressing in the preface to the Tractatus.  His work is an attempt to ‘draw a limit to thinking, or rather—not to thinking, but to the expression of thoughts’.  The distinction between saying and showing is Wittgenstein’s solution to the problem of critical philosophy: to draw the limits on what can be meaningfully said[17], ‘and what lies on the other side of the limit will be simply nonsense.’

Wittgenstein only uses the word ‘transcendental’ twice in the Tractatus.  Once (in 6.421) to consign ethics to the transcendental, and once with regard to logic:

6.13 Logic is not a body of doctrine, but a mirror-image of the world.

Logic is transcendental.

Despite this frugal use of the word, it is enlightening to compare the ‘transcendental’ in Wittgenstein with that which ‘shows itself’ as opposed to that which can be said.  It is clear, especially from the Notebooks, that Wittgenstein was greatly influenced by Schopenhauer, but it could not be said that the final form of the Tractatus has any straightforward correspondence to the idealism of Schopenhauer.  The most striking aspect of Wittgenstein’s approach to philosophy is his concern with the limits of sense, and this is closely related to the problem that Kant addressed in The Critique of Pure Reason.  Kant used his transcendental philosophy to delineate what could be known.  Wittgenstein was interested in the limits of what could be said (by which he seems to mean what could be described or asserted), and his transcendental philosophy is an attempt to draw these limits.  But therein lies a problem, for to see a boundary one must see the other side, or at least confront its impenetrability.  What marks the boundary of meaning is not something that can be, strictly speaking, described.  Thus the transcendental philosophy of the Tractatus is inherently ‘mystical’:

6.522 There are, indeed, things that cannot be put into words.  They make themselves manifest.  They are what is mystical.

For Wittgenstein, the transcendental cannot be expressed, but shows itself — it makes itself manifest in the ‘unalterable form’ (2.022 - 2.023) of the world and language, but any attempt to express explicitly [sagen] what that form is, necessarily results in nonsense.  The propositions one tried to produce in such an enterprise would not be propositions at all, for they would be nonsensical, and genuine proposition cannot fail to have sense.  This enterprise is, in fact, what most of philosophy consists in, and consequently its propositions and questions more often than not turn out to be not propositions or questions at all: they are nonsense and the questions neither have answers nor require them.[18]

The connection between what can only be shown (the form of logic, language and the world that makes itself manifest) and the transcendental can best be understood if we bear in mind the emphasis Wittgenstein placed on the philosophy of language.  Just as Kant sought to uncover the subjective conditions of the possibility of knowledge and experience, Wittgenstein’s early philosophy sought to discover the conditions for the possibility of language and thought.  Wittgenstein thought that it was precisely because these conditions were always presupposed in language that they could not themselves be represented:

4.12 Propositions can represent the whole of reality, but they cannot represent what they must have in common with reality in order to be able to represent it—logical form.

In order to be able to represent logical form, we should have to be able to station ourselves with propositions somewhere outside logic, that is to say outside the world.

In order to consider the conditions of language and the world, one must take up the transcendental perspective that is outside logic and the world.  But in taking up this perspective, we necessarily leave behind the conditions that give our thoughts meaning.  The best we can do is to allow those conditions to make themselves manifest as a part of our ordinary thoughts:

4.121 Propositions cannot represent logical form: it is mirrored in them.

What finds its reflection in language, language cannot represent.

What expresses itself in language, we cannot express by means of language.

Propositions show the logical form of reality.

They display it.

Now, despite the claim that logical form cannot be stated by means of propositions, this is exactly what Wittgenstein attempted to do from the first page of the Tractatus.  He built a picture of language and the world that makes clear how both can share their logical form.  This picture building is presumably an attempt by Wittgenstein, not to say what logical form is, but to bring the logical form of language and the world into focus for the reader: to show us the logical form of language and the world.

2.4 The Idealism of the Tractatus: Logic and the World

We now have a certain picture of the relation between language and the world in terms of a common logical form.  It is worth pondering for a moment just how tight a connection between the world and language Wittgenstein has proposed.  ‘The world is all that is the case’[19] and nothing more[20].  And what is the case can be represented:

2.18 What any picture, of whatever form, must have in common with reality, in order to be able to depict it—correctly or incorrectly—in any way at all, is logical form, i.e. the form of reality.

2.181 A picture whose pictorial form is logical form is called a logical picture.

2.182 Every picture is at the same time a logical one.  (On the other hand, not every picture is, for example, a spatial one.)

2.19 Logical pictures can depict the world.

Moreover, we cannot even conceive of a state of affairs that cannot be pictured, for to conceive of a state of affairs is just to picture it:

3 A logical picture of facts is a thought.

3.001 ‘A state of affairs is thinkable’: what this means is that we can picture it to ourselves.

And language is a pervasive phenomenon.  To think of the world at all, to project a proposition, is to use ‘language’ in some sense.  All thoughts (Gedanke) also have logical form.

4 A thought is a proposition with a sense.

4.001 The totality of propositions is language.

Now it is becoming clear why I thought that thinking and language were the same.  For thinking is a kind of language.  For a thought too is, of course, a logical picture of the proposition, and therefore it just is a kind of proposition.  (Notebooks 11.9.16)

Now, I mentioned at the beginning of this chapter that Wittgenstein’s picture theory appeared to have much in common with a correspondence theory of truth, but several other elements are now in place that distinguish it from this kind of realism.

The realist insists that propositions must correspond to an independent world, and the idealist contends that we have no access to that which is mind independent.  Limited by our own consciousness, we cannot make the distinction between the ideal and the real.  The transcendental idealist, however, claims that this distinction can be made, but one must bear in mind on what level it is being made.  At the empirical level it is simply the distinction between minds and mental content on the one hand, and the rest of the real world those contents correspond to on the other.  At the transcendental level, however, the distinction is between the world that we experience and represent, and which is therefore necessarily subject to the conditions of experience and representation, and the world considered (as far as it can be) as independent of those conditions.  It is at the transcendental level of analysis that these conditions are considered as forming the necessary structure of both knowledge and the world.  Wittgenstein’s approach in the Tractatus can be seen as a kind of transcendental idealism.  The world, under this conception, is the phenomenal world, where to be part of the phenomenal world is to have logical form.  It is the world considered as subject to the conditions of description and representation – the world that has logical form.  It is therefore this logical form that limits the world.

2.5 The Solipsism of the Tractatus

By now, the meaning of 5.6 and 5.61 should be clear:

5.6 The limits of my language mean the limits of my world.

5.61 Logic pervades the world: the limits of the world are also its limits.

So we cannot say in logic, ‘The world has this in it, and this, but not that.’

For that would appear to presuppose that we were excluding certain possibilities, and this cannot be the case, since it would require that logic should go beyond the limits of the world; for only in that way could it view those limits from the other side as well.

We cannot think what we cannot think; so we what we cannot think we cannot say either.

The limits of the world are it logical limits, which are given by its logical form.  According to the metaphysics under discussion, nothing in the world, or in language, can fail to have logical form.  That form therefore marks the limits of both language and the world.  Thus the first line of 5.61 can be taken to be an expression of the transcendental idealism already evident in the metaphysics of the Tractatus[21].  What follows can then be seen as a rejection of the idea that one can talk about the transcendentally real: we cannot say, in a language that is based on logical form, what there is that is independent of logical form.  The transcendentally real falls outside of logic, language and the world.

And of course 5.6 follows from 5.61: the limits of both my language and my world are given by their logical form.  But why introduce the first person at all here?  Why speak of my world and my language?

Consider again the picture of the relationship between language and the world in terms of logical form.  Where does the subject fit into that picture?  One answer, suggested by 5.542, is that the subject contains the names that are configured in such-and-such a way.  The empirical subject must be a part of the world, and therefore consist of objects standing in relation to one another.  (And some of the facts of which it consists will constitute propositions).  But this would, at best, only be the empirical subject — which is not the metaphysical subject discussed in §5.6ff.  (In any case, 5.5421 indicates that this empirical ‘subject’ is not considered a subject (soul) at all by Wittgenstein, on the basis that it is composite).  

What then is the metaphysical subject discussed here?  The possibility of language is the possibility of thought about the world.  But what does this possibility consist in?  It consists in the world and language having logical form.  The subject is introduced as a kind of transcendental subject[22]: it is the boundary of the world and language in that it is presupposed by the possibility of world and language.  This is what is revealed in the fact that the world is my world and language is my language.

Wittgenstein eventually admits (in 6.54) that his propositions are nonsensical, stating that their value lies not in their literal meaning, but in their power to elucidate.  The picture of language and the world that emerges from his philosophy is not one that can be described, but rather one that makes itself manifest to someone who follows Wittgenstein’s thoughts.  Wittgenstein is trying to show us that which cannot be said.  But whose language emerges as connected with the world in this way?  Whose language shares its limits with the substance of the world?  If this picture is made manifest to me, by my reflection on my language and my world, then there can be only one answer to that question.  It is my language.  Furthermore, I am stuck within the bounds of this language: anything that counts as a proposition for me will be a proposition in my language.

This is how Wittgenstein first explains how solipsism is made manifest in 5.62:

The world is my world: this is manifest in the fact that the limits of language (of that language which alone I understand) mean the limits of my world.

“of that language which alone I understand” is a translation of “der Sprache, die allein ich verstehe”.  Both in the German and in the translation given by Pears and McGuinness, this is an ambiguous description.  One obvious translation emphasises the ‘I’ to produce a statement of crude solipsism: no one else can understand my language and my thoughts — they are ‘private’.  On this reading, solipsism is supposed to follow from this privacy of language.  I will call this reading the ‘privacy of language’ reading.  It has something in its favour, since one might think that the privacy of language is implied by Wittgenstein’s picture theory.  While all language shares the same logical form, my language is not the same as your language.  The analysed form of sentences that attribute mental predicates to me will presumably be different to those attributing mental predicates to others.  For example, there will be an asymmetry in my language between the proposition that says, “I am in pain” — or better, “There is pain” — and the proposition that you are in pain.  This is the asymmetry that is emphasised in Wittgenstein’s later work[23].  The point is that my language has a certain ‘perspectival’ relation to the world.  It is not clear, however, to what extent this perspectival nature implies privacy of language.  The strongest kind that could possibly be attributed to Wittgenstein would be if all the objects of the world were private to a particular subject, as the sense data of acquaintance are private in Russell’s version of logical atomism[24].  Indeed, a comparison with Russell’s logical atomism would seem to provide the most persuasive evidence that what is at issue in this passage is precisely the question of privacy.  Russell’s objects are those that are named with the word ‘that’.  That can be, as it were, pointed to inwardly, and Russell admits that this inward acquaintance is private.  While the Objects of the Tractatus are not introduced in the same fashion, but are rather required on transcendental grounds, Wittgenstein did seem to believe that that it was possible to examine them as private objects:  he thought that that we learn how to form propositions from names by examining the Objects they corresponded to[25].  While the view in question does not strictly imply privacy, one would suspect the early Wittgenstein of holding that this examination goes on in private from the extent of his attack on private definition in his later work, which to a great extent is a critique of his earlier views.

Be this as it may, the privacy reading of §5.6ff is flawed in a crucial respect: that the kind of privacy in question does not lead straightforwardly to solipsism.  To get there requires the addition of a kind of scepticism that one would rather associate with empirical (rather than transcendental) solipsism or idealism.  Having distinguished my language from everyone else, I must then use an epistemological scepticism of the kind used by Descartes and Hume[26]: I deny the existence of other subjects on the basis that I have no evidence of them.  Not only does this seem to be far removed from Wittgenstein’s thoughts in §5.6, it is not even consistent to use the notion of privacy here either.  For while I have distinguished my experience on the basis of privacy, that concept is undermined by the scepticism that follows from it.  There can be no privacy where there are no other subjects to ‘not get a look in’, so to speak.

Perhaps it is more in line with the spirit of §5.6 if we disambiguate the description in the other way, by binding the ‘allein’ (‘alone’) to the language.  As the later version of the Ogden translation puts it, it is “the language I understand”[27].  That is, the only language I understand is my language, and I have no way to climb outside it, and view the world from some point external to my thoughts.  My language is the only language I have, and it represents the only world I can think about: my world.

Now, this could be seen as another way of endorsing a perspectival view of language.  One might say that this is required to emphasise the uniqueness of my language, and thus justify the introduction of the first person.  Furthermore, one might want to express this in terms of the privacy of my language, thus blurring the distinction we made between the two interpretations of 5.62.  But such an interpretation must be treated with caution: the solipsism of §5.6 does not individuate my language from many.  It does not pick out my world from the worlds of other subjects.  My language is the only language I have, and it holds me captive inside my world.

The most important distinction between the private language interpretation and the kind of interpretation that I would like to advocate is that former uses the notion of privacy as an antecedent in the argument to solipsism.  This leaves it mysterious why Wittgenstein would want to argue for such an esoteric position, and indeed why he introduces the subject at all.  On the other hand it is clear that, since Wittgenstein does introduce a metaphysical subject, the question of subjectivity is at issue.  Perhaps the issue of privacy should be seen, not as explanans, but as explandum.

After all, Wittgenstein’s idealism has a certain problem.  There does seem to be something inherently private about my experience.  It is not so much that my language is inherently private, as that my language is inherently public.  If the limits of language are the limits of the world, then what of the facts that seem to fall outside of language?  What about the facts that I cannot describe, namely the way thing seem to me?  As David Bell has put it, the whole of subjectivity seems to fall outside of the machinery of objectivity[28].  But of course Wittgenstein has an obvious answer to this problem.  Subjectivity is part of that which shows itself.  That which shows itself but cannot be said constitute the subjective conditions of language and the world.

Logical form is made manifest in the fact that there is a world at all.  That there is a world at all is connected with the fact that I live.  Wittgenstein expressed this at one point in the Notebooks by noting that the subject is not a part of the world but a presupposition of its existence (NB 2.8.16)[29].  He expresses it in the Tractatus by saying, “The world and life are one” (5.621), “I am my world” (5.63) and by saying that the subject is not a part of the world, but a limit of it (5.632).  And if the subject is a necessary condition of the world, it cannot be a contingent part of that world.  (And if it was, we would not have transcendental solipsism, but empirical solipsism - things would depend on my perceiving them).  This, I take it, is the main point of the metaphor of the eye in 5.633, and at least partly explains the remarks of 5.634:

5.634 This is connected with the fact that no part of our experience is at the same time a priori.

Whatever we see could be other than it is.

Whatever we can describe at all could be other than it is.

There is no a priori order of things.

Wittgenstein held the general principle that if something is known a priori, it cannot also be part of the world.  A priori propositions are not meaningful in that they say nothing about the world.  The only propositions with sense are empirical[30].  Nevertheless, thinking of the subject as that which is presupposed by the possibility of representation and the world is a kind of transcendental solipsism: A transcendental idealism considered for my world and my language.  This form of expression fits with Wittgenstein’s own report of the development of his ideas:

This is the way I have travelled: Idealism singles men out from the world as unique, solipsism single me alone out, and at last I see that I too belong with the rest of the world, and so on the one side nothing is left over, and on the other side, as unique, the world.  In this way idealism leads to realism if it is strictly thought out.  (NB15.10.16)

Transcendental Idealism leads to solipsism, but not a solipsism that can be expressed, but one that shows itself.  The transcendental cannot be said, and yet the world consists only of what can be said.  (The world is the totality of facts, and facts can be described).  So we are left only with the empirical facts, and hence a kind of empirical realism.

Now, this interpretation still leaves us with a problem, for it is still unclear in what sense the subject can be considered as a ‘limit’ of the world.  In 5.6 — 5.62, when Wittgenstein refers to the limits of language and logic also being the limits of the world, he must surely be referring to the limits he discusses elsewhere in the Tractatus: limits that are ‘logical’ limits in quite a narrow sense.  Contradictions, propositions with sense, and tautologies correspond to the notions of impossibility, possibility and (analytic) necessity.  The limit of the world is given by what is possible, and that, according to Wittgenstein, coincides with what can be pictured.  Since contradictions and tautologies picture nothing (any pictures—propositions—contained within them ‘cancel out’), they say nothing about the world.  But they show the logical form of the world by showing something about what is essential to picturing it.

By 5.632, however, Wittgenstein refers to the subject as ‘a limit of the world’.  Again in 5.641 Wittgenstein talks of ‘the metaphysical subject—the limit, not a part of the world.’  I have claimed that this is because the subject must be presupposed by the possibility of the world.  But how does this idea fit with Wittgenstein’s conception of logic?  Moreover, some of the statements of §5.6ff do not seem, at first sight, to be consistent.  How can the subject (me) be a limit of the world, while at the same time ‘I am my world’ and ‘The world is my world’ (from which we can derive ‘I am the world’)?

The first point to note is that the word ‘limits’ [‘Grenzen’] can be given a more general meaning that the strict logical limits of tautology and contradiction.  The limits of the world include the logical form of the objects, for example.  In general, the limits of the world are the limits of language: that which can be shown but not said.  For this showing takes the form of deriving the conceptually necessary conditions of language (and the world that can be described by means of this language).  Being necessary conditions of that world, they are neither a part of it, nor independent of it.  They manifest its limits.  Thus, all the transcendental concepts in the Tractatus, including ‘objects’, ‘logic’, and even ‘language’ and ‘the world’ are limits of language and the world.  The world in its entirety is not something that can (strictly speaking) be referred to, but lies at the limit of what can be said.[31] 

The second point is that when Wittgenstein relates solipsism to his discussion of the limits of language he is working together two separate lines of thought from his notebooks.  In fact, the Notebooks contain a huge variety of ideas, which Wittgenstein tries to draw together, linking each one back to his ideas on logic.  This is as one would expect from the ambitious nature of the project.  Wittgenstein was trying to settle all of the questions of philosophy once and for all.  The sources of the various ideas he looks at are diverse, but he constantly tried to unify them in a single scheme of ideas, so that they can all be ultimately settled in a similar fashion.  For example, the ideas on solipsism can be found in the coded sections of his notebooks — that is to say his more private diary — long before they appear in his philosophical notes.  So at least one motivation for taking solipsism seriously seems to have been the loneliness and isolation he felt while serving on the eastern front.  At one point in his coded diary he remarks on these solipsistic ideas that, ‘Oddly enough,’ he could not make the connection with his ‘mathematical modes of thought.’[32]

But in the next few weeks we find that connection being forged in the philosophical notes in the ideas that eventually formed §5.6.  Wittgenstein, I believe, linked his ideas in the following way: language and the world make manifest a picture of their own form.  This form implies a kind of solipsism, and this explains the philosophical tendency to solipsism: the fact that solipsism is also made manifest in our ordinary experience.  The fact that everything seems to be, at one and the same time, both part of my subjectivity, and part of the world.

The connection between these ideas can best be made by stressing the sense in which the solipsism of the Tractatus also concerns subjectivity.  When Wittgenstein says that what solipsism means is quite correct, only it cannot be said, but makes itself manifest, he is inviting us to reflect on the world in a certain way.  Transcendental reflection here becomes at once both the reflection of the necessary structure that the world and language must share, and reflection on the world as experience: ‘The microcosm’.  And when we reflect on the world in this way, what we want to say is this: that the world is my world.  It is not that it is my world as opposed to yours.  That is not the point.  The point is that I am struck by a certain limitation on the world: that everything in it is in some sense mine.  When I reflect on the world as experience, I only have my experience to reflect on.  But of course, this way of expressing the matter will not do.  On Wittgenstein’s theory of meaning, this kind of reflection cannot be expressed at all.  It is part of what makes itself manifest.  And that this particular way of trying to express it will not do is shown by the fact that, once I have started to reflect on the world in this way, there can be no me in this experience to own it. 

The expression ‘I am my world’ will not capture this form of reflection either.  The ‘I’ of this expression refers to nothing in the world.  Again, on Wittgenstein’s theory of meaning, this makes the expression meaningless.  Furthermore, we cannot refer to ‘the world’ either.  Nevertheless, in following this meaningless line of thought we come up against ‘something’ that shows itself but cannot be said.  We are confronted by it.  And that we are confronted with it explains why a certain philosophical line of thought is tempting.

Now the more analytically minded readers will find some of this reasoning unsatisfactory.  Compared to the tight logical reasoning of much of the Tractatus, the comments in §5.6 have struck many readers as incomprehensible, or unnecessarily mystical[33].  And I must admit that, even if the interpretation given here is correct, it leaves a great deal mysterious.  For example, for all the flesh I have tried to put around the idea that the limits of the world are the limits of my language, it remains mysterious why Wittgenstein felt entitled to introduce a metaphysical subject at all here[34].  Ultimately one is tempted to resign oneself to the idea that Wittgenstein simply made a mistake.  Given that the metaphysical subject and solipsism seem to be the target of a great deal of Wittgenstein’s later critique of his early work, it is clear that he came to think this himself.  (But by the same token, it shows that he continued to think that the temptation to think that way was important).  We will return to this matter in Part 3 of this thesis, when we discuss quietism with respect to subjectivity.  For now we will get by with a brief discussion of that element of the Tractatus that I think provides the greatest clue to understanding these matters, and indeed his approach to metaphysics throughout his work.  This is the fact that Wittgenstein eventually concedes: that one can, strictly speaking, say nothing on these matters.

2.6 The Quietism of the Tractatus

We have seen that the transcendental idealism of the Tractatus combines both a conceptual idealism and a conceptual realism in its metaphysics.  It presents a conceptual idealism in that the world is limited by the conditions of representation.  The world is all that is the case, and the possible facts correspond to that which can be pictured.  But in order to discuss the transcendental conditions of representation and the world, Wittgenstein had to discuss those conditions themselves, such as the necessary existence of objects, which could not be themselves subject to those conditions.  In order to combine this realism with idealism, Wittgenstein made a distinction between that which can be said and that which makes itself manifest.  For the most part the subject matter of Wittgenstein’s metaphysical statements is the latter, that which can only be shown.  So these propositions themselves are nonsensical, since the subject matter of any meaningful proposition can only be that which can be said.  Their value, according to Wittgenstein, lies in their elucidatory power:

6.54 My propositions serve as elucidations in the following way: anyone who understands me eventually recognises them as nonsensical, when he has used them—as steps—to climb up beyond them.  (He must, so to speak, throw away the ladder after he has climbed up it.)

He must transcend these propositions, and then he will see the world aright.

Seeing the world aright, the reader is supposed to be cured of the temptation to utter nonsense.  The final position we are left with is neither idealist nor realist, since neither of these positions can be meaningfully stated.  The final meta-philosophical outcome is quietism: “Whereof one cannot speak, thereof one must remain silent.”[35]

We are still left, of course, with a feeling of discomfort: a feeling that a trick has been pulled.  There is a contradiction, of which, I think, Wittgenstein was perfectly aware.  To say that we have come up against something that cannot be said is to contradict oneself.  Something that cannot be said?  But have I just not said it, by referring to it as a ‘something’?  And there is no way to re-express the matter so that the ‘something’ drops out, and the contradiction with it[36].  Sometimes this contradiction appears to be a tautology: ‘We cannot think what we cannot think’ (5.61).  But the contradiction always remains, even in the final proposition of the Tractatus, when Wittgenstein uses the phrase ‘Whereof we cannot speak’.  So right to the end, even after Wittgenstein has ‘pulled up the ladder,’ he leaves us with a contradiction.  But I think it is a contradiction that bears reflection.  For that very contradiction makes manifest the limits of what can be said.